In mk_MK.ISO8859-5 locale on Solaris 8, the getDisplayLanguage() routine
returns a String that cannot be printed. Though the font used in Bugtraq distorts
a lot, the printed characters look like "MaKeAo?CK", followed by a backwards "N".
The question mark "?" locates an incorrect character in the value returned by
getDisplayLanguage().

import java.util.Locale;
public class Hello {
public static void main(String[] argv) {
    String localeString =
        Locale.getDefault().toString() + "/" +
        Locale.getDefault().getDisplayLanguage();
        System.out.println(localeString);
}
}

http://www.cs.earlham.edu/~dusko/InfoMak/language/ contains a gif that shows
that the return value should look like. The place marked by the "?" should
display a capital letter H.

allan.jacobs@Eng 2000-03-21

I logged into the mk_MK.ISO8859-5 locale on a Solaris 8 machine. In a X-window,
I switched the input language from English to Macedonian (left click in the textfield
at the lower left and chooser the other language). The correct return value
for getDisplayLanguage can be obtained by typing "vfrtljycrb" on an English
language keyboard. Using cut-and-paste, these characters were put into a file.
The file was dumped with "od". They all fit into bytes and their hex values are
dc d0 da d5 d4 de dd e1 da d8. Hex dd is 221 decimal.
M a K e A o H C K N

allan.jacobs@Eng 2000-03-21